Data Element

Information Gathering

2010-05-05T09:43:00.003-06:00

My way to view business requirements gathering for data warehousing and business intelligence projects is as a process to collect or gather information needs. Therefore, I see the process more focused on finding those business questions that require the right answer based on a single version of facts (i.e. the data warehouse).
These findings of information needs have to be encompassed with verification of data availability, which implies more that knowing that there is one or more sources of data containing records that can be utilized to answer the business question. It is important to confirm access to the data, frequency of the refresh, feasibility of the data integration, and many other technical and business considerations.
I indicate business considerations, as sometimes if not always, getting access to the right data involves political decisions.
After confirming that data is or can be available in the data warehouse, then we kind of cover the what (data) and how often (frequency) the answers will be ready to be used. This is not enough, as we have to deliver this information at the right time and in the right format. It is then when we start moving towards the business intelligence part of the solution. So far, we have travesed the data integration including ETL-ing and the data warehousing components.
It is important to filter out information wants from actual information needs. During the exercise, you will realize that decision makers will describe those information needs in different formats or presentations. You have the task of extracting measures and dimensions from each statement. In addition, you have to identify cross-functional requirements involving more than one business unit.
Business questions relate to one or more operational systems. A matrix is helpful to cross-validate and organize systems, measures, dimensions, and hierarchies.
PENDING: Business goals, business objectives, key contacts, etc.

Analytic Applications

2010-04-05T14:40:00.002-06:00

What is an analytic application?
This is my outline to respond this question:
- Business Analytics vs. Business Intelligence
- Architecture of Analytic Applications
-- 3-Tier Architecture
-- Service-Oriented Architecture
- Reporting
- Performance Dashboards
- Other forms of Information consumption
- Analytic Applications

Data Profiling

2010-03-30T16:25:00.003-06:00

What are the benefits of Data Profiling? A data profiling tool has been added to SQL Server Integration Services 2008. There are some tabs available in the data source view in SQL Server Analysis Services that are helpful for data understanding in relation to data profiling.

Is data profiling more than statistics about data? how does data profiling contributes to better data management? This is my initial paragraph for a post that is intended to address data profiling.

Slowly Changing Dimensions

2010-02-22T07:48:00.002-07:00

I have been asked about What are the benefits of using Slowly Changing Dimension Type 4? My first answer was that SCD Type 4 are justified for data records that are frequently updated but only the current version is relevant for analysis. In addition, history for all changes apply to those records is used for Knowledge Discovery in Databases (KDD).
Under this scenario, a SCD Type 4 does not impact performance as it only increase size when new records are added to the dimension table. However, all changes are still available in the history table for this dimension.
Now, Are all changes going to trigger a new record in the history dimension table? I do not think so. During the same discussion that initiate this post, it was clear that there are columns that are not relevant for the business and SDC Type 1 is suggested. In other cases, there are columns that will require only previous and current version. SCD Type 3 is suggested there.
When history is not needed given the amount of expected changes that will trigger new records, the size of the dimension table, the relevance of tracking those changes during OLAP, etc. the recommendation is to utilize SCD Type 2.
What about SCD Type 6? Well, as discussed above, combining different SCD Types for a particular dimension table seems to be a necessary thing.
Again, all depends on business information needs and that suitable model that will satisfy such needs.

Data Integration and Business Intelligence Process

2010-02-03T22:46:00.006-07:00

The Data Integration and Business Intelligence process consists of five steps: Gather, Collect, Organize and Select, Synthesize, and Distribute.
Data is gathered from internal and external sources. Internal sources include such systems as ERP systems, Legacy systems, etc. Based on information needs, the data gathering step serves to identify data sources, which includes data servers, database schemas, user accounts, roles and privileges to access data, data availability, frequency of data refresh processes, data dictionaries, metadata, data owners, data stewards, and other relevant information that will facilitate the data collection. Data gathering tasks are supported by robust tools such as Informatica DataExplorer.
Data Extraction, cleansing, transformation and loading into central repositories (or Data Hubs) is carried out utilizing data integration tools. A critical aspect in data extraction is managing sensitive information using processes that must comply with information acts (e.g. FOIP, HIA). Personally Identifiable Information can be protected using data encryption techniques. For data validation, it is critical to have the business rules at hand to validate the data domain, data definitions, and data constraints from the business perspective. On the other hand, the extraction tool supports validation of different types of data integrity such as domain integrity, entity integrity, referential integrity, and user-defined integrity. Missing values found during data extraction, cleansing and validation can be extrapolated using data mining techniques. As for data loading, there are different approaches. The one that is highly recommended is the populated a staging database, perform complex processing in a dedicate server (hosting the data engine) and then move data elements to the corresponding schemas in the data hub. An example of complex processing is management of unstructured data.
Once data in collected in a Data Hub, further data profiling processing and data matching can be applied. In addition, Change Data Capture software is required to determine and track data changes made to the participating data sources in the data integration process. Record matching, duplicates detection, and other related operations to record linkage are supported by advanced data integration tools. Matching records or eliminating duplicates in data coming from heterogeneous sources is one of the most difficult data integration challenges. Some scenarios will require human intervention and a powerful data integration tool.
The Data Hub includes a staging area containing one or more staging databases for pre-processing tasks, such as cleaning, normalization, transformation, feature extraction, and selection. Pre-processing tasks can be utilized to select data for populating a dimension table in a data mart or to determine the data elements to be used in the hierarchies for the same dimension.
Once data is pre-processed, data is organized and selected. The data integration tools are then used again to extract, transform and load data into a enterprise data warehouse, data marts, or other target systems.
Data stored in the enterprise data warehouse or data marts can be synthesized to facilitate analysis by slicing, dicing and drilling through cubes based on fact tables, dimensions and hierarchies. Data mining techniques, such as clustering can be used to carry out unsupervised data partitioning. On the other hand, the association rules data mining technique can be used to find unsupervised (without human interaction) data matching in thousands or millions of records. Data mining also can be applied to discover patterns in data as part of a knowledge discovery process.
Information is distributed through different channels. It is important to emphasize that data for reporting is pulled out from the Data Hub, the enterprise data warehouse, and analytical views.
Information delivery is not limited to pre-defined or "canned" reports; it is also available through other data access services such as standard database connectivity (JDBC/ODBC) and web services. Information delivery is enhanced by using advanced reporting tools, such as report generation using multidimensional expressions (MDX) to pulling out information from data cubes.
Information is distributed via email, web applications, desktop applications, portal services. Furthermore, information can be presented in dashboards and balanced scorecards to enable performance management. Performance management tools allow business users to monitor, analyze and plan their business as well as make well-informed decisions faster. Dashboards are highly summarized, often graphical representations of the state of the business. The values on a dashboard report are often key performance indicators (KPIs) for an organization. A dashboard normally includes comparisons to targets.

Wikies and Social Networks for Building Strategies

2010-01-29T07:34:00.003-07:00

Data Strategies, BI Strategies, etc. are built hiring an external consulting company that brings best practices and expertise. Nothing wrong with that approach. However, the consultant assigned to help the organization and put together the strategy, which is normally a project outcome in the form of a document.
This document ends in a shelf. The organizations claims to have a "strategy" and points to the document to support the statement. Well, one of the reasons, in my opinion, is that the organization understood a collective of human beings does not see the document as theirs. In order to change that mindset, personnel should be heavily involved in defining, organizing, building and maintaining the strategy. Then, they will not only claim that they have a strategy but for sure they will mean that it is theirs.
What should be the approach? My suggestion is to create a wiki and utilize the power of social networks to create and maintain the strategy. In this case, the social network is within the boundaries of the organization, though there is not reason to allow not external participant.
The consultant might contribute indicating the framework that will be used to build the strategy. After a period of time and when enough contributions have been made (this part will require a good awareness campaign, motivation and follow ups), the consultant will join the efforts by bringing those best practices and expertise and use them to help the organization structure and get the "document" that represents the strategy.
The document must be a live document that is transformed with the organization. It is not a static object located somewhere in a shelf.
This is the way I envision how wikies and social networks can be used for building strategies.

BI for more Effective IT - Business Model

2010-01-19T13:36:00.002-07:00

I work for an IT company, which has been in the market for 20+ years. I have been considering focusing on IT for the Business Intelligence Practice. Now, the question is, Do I have a business model to present? The answer to the question is the main motive of this blog.

The Business Model
1. Value Proposition
- A description of the customer problem
- The product that addresses the problem
- and the value of the product form the customer's perspective
2. Market Segment
- The group of customers to target, recognizing that different market segments have different needs. Sometimes the potential of an innovation is unlocked only when a different market segment is targeted.
3. Value Chain Structure
- The firm's position and activities in the value chain and how the firm will capture part of the value that it creates in the chain
4. Revenue Generation and Margins
- How revenue is generated (sales, leasing, subscription, support, etc.) the cost structure, and target profit margins
5. Position in Value Network
- Identification of competitors, complementors, and any network effects that can be utilized to deliver more value to the customer
6. Competitive Strategy
- How the company will attempt to develop a sustainable competitive advantage, for example, by means of a cost, differentiation, or niche strategy.

Information Provider - Knowing what they would like to know.

2010-01-14T10:23:00.004-07:00

I am working on how to utilize knowledge discovery to benefit member- or subscriber-based organizations.
Is it not only about what you don't know about your customer but also what you should know about your customer interests and preferences.
Customer interests and preferences could have been identified already, but how is the organization using that knowledge to provide information to its customers. Sometimes organizations would like to predict sales, market trends, etc. What about anticipating customer's information needs and providing them with timely and accurate information so a customer can make an well-informed decision.
In other words, knowing what your customer would like to know and presenting information in a simplify manner.
The idea is not to bombard your customer with facts and offers, but to send the right information at the right time. There is the value of Business Intelligence.
Now, how to identify customer's information needs? how to identify the "right time" to send the "right information"? how to determine that it is enough information for now? how to get feedback from the customer and improve information delivery mechanisms?
Good questions that experience (people) and technology can help respond. The missing part of the puzzle is the process to collect data, analyze data, transform data into information that will satisfy customer's information needs, then make information actionable.
Brainstorming for an exercise that I am doing with a friend of mine, I thought the following: I am a Alberta Motor Association (AMA) member of have some consumer habits, e.g. fill up my car every week and use to look for gas stations with AMA sign to earn AMA dollars for purchasing there. On the other hand, I am a GPS user that use to look for restaurants, etc close to my current location.
It would be good to get recommendations of the closer stores with AMA agreement that will help save some money through discounts for AMA members only or to earn a couple of AMA dollars. If this information is available in my cell phone, ever better.
Now, there are AMA services available to members and non-members such as road conditions. This information can be enhanced to members if they enter their membership number and the "system" can use data mining to find associations, classifications, etc that will create a more personalized service. All this as an on-line service via Internet delivered by AMA.
Now, what is the framework needed for implementing this kind of on-line services? how can data mining contribute to enhance those services? is cloud computing an option now (or later) for deploying an application for on-line information services? is Software as a Service (SaaS ) an alternative to optimize IT resources in this line of services? how can open data (as the one the City of Edmonton is making available, see http://data.edmonton.ca) be utilized to complement these information services? Again, what is the framework required for supporting these information services?
Like the exercise, but soon I have to start answering so many questions.
In summary, organizations can combine their data and open data to create a source for data analysis and generation of information assets. The process should be driven by customer's information needs. Finding patterns in the way customers search or seek for information can be used to deliver the right information at the right time to the right customer.
Customers will welcome this appealing information that will help them make well-informed decisions. This could derive into customer satisfaction, customer retention, cross-sales, etc. It is beneficial to the customer and to the organization. Like win-win approaches.

Total Information Outsourcing

2010-01-13T16:53:00.003-07:00

According to the article Cloud Computing and SaaS: New Data Mining tools for the IRS? (Jean-Paul Smets-Solanes & Rogerio Atem Carvalho, published in the Enterprise Management Quarterly's Web site, a business can implement Total Information Outsourcing (TIO) and still protect their data. In the article, TIO is defined as follows: "the concept of Total Information Outsourcing consist of implementing the whole information system of an organization through Web based services using technologies such as Cloud Computing or Software as a Service (SaaS)."
There are key questions asked in this article, such as: Can an organization get back its corporate data? and Can the organization get a copy of the software use to process it data? According to the authors, the answers are no to both questions. Scary, isn't it? But they provide a way out, named data portability. At least for the access to the data. They also mention that the Foundation for Free Information Infrastructure has created a group to explore ways to minimize the risks for companies wishing to adopt TIO.
It is encouraging to see progress in this area. At the same time, it is valuable to explore options and contribute to this area.

Open Data and Information Acts

2010-01-13T16:43:00.003-07:00

What is the implication of using open data regarding information acts? Freedom of Information and Protection Privacy (FOIP), Health Information Act (HIA) and others are regulatory mechanisms to protect data. How is open data going to change the way organizations are facilitating access to their data? Furthermore, how are organizations going to protect the data they are stewards of?
This is an interesting subject for me and it entails data security, data encryption, data protection, data dissemination, etc.
I would like to explore more in detail business and technical considerations on this matter. Let's see what can I gather from now on.

Framework and Infrastructure for Embedded Data Mining

2010-01-13T07:40:00.003-07:00

After discussing with a friend of mine, some ideas for utilizing data mining, it came to the conversation the importance to establish a foundation for a solution that will use data mining capabilities.
My suggestion was to define an infrastructure that will accommodate to future solution. Then, he sent me an article that talks about a framework for data mining in the cloud. Interesting thing.
In order to establish such a foundation, it is necessary to define a framework, the infrastructure and then start working on automation that will facilitate embedding data mining capabilities.
I envision this as an option to have a data mining engine working as a search engine does. Or similar to the "functions" in Excel, where the user with a brief description of the inputs, outputs and examples utilizes functions that are complex and powerful for statistical analysis. The user doesn't know details on how the function works, or the theory behind. It knows the function works, it does what s/he is expecting, and that's what matters to him/her.
There is a lot of progress on that regards and there are available adds-on or plug-ins that enable software products with data mining capabilities.
A review of what have been done and is available, what is in progress and what should be done targeting technologies that will be available in the coming 5-10 years is an interesting subject. I am more interested in the last one.
Let's envision the framework and infrastructure for embedded data mining that will use or leverage current things such as Web 2.0, Web services, Social networks, Cloud computing, Mobile computing, Advanced visualization, etc.

Data Mining a Tool for Data Understanding

2010-01-13T07:33:00.004-07:00

I have found some growing interest in making Data Mining accessible to the masses. It is great. Furthermore, I consider that data mining has been out there for "the masses" for quite a while. I am thinking on Amazon.com and the typical example used for Data Mining presenters about "People that have read this book have also read these other books". The market-basket analysis has been available for some years, as far as I know and have utilized it.
Now, what would be best, to facilitate data mining tools through user-friendly interfaces or to embedded data mining capabilities and present results that will support the data understanding process.
At this time, I am more inclined to the second option. Let's discuss this options in more detail.

Data Understanding - Principles & Practices

2010-01-12T10:11:00.003-07:00

What are the Principles of Data Understanding?
Principles never change.

What are the Practices of Data Understanding?
Practices are based on principles and can be adjusted to circumstances, e.g. size of the organization, maturity in the market, etc.

What are the current best practices for Data Understanding?
Best practices can be either internal or external. External best practices are accepted by an industry or by a group. Internal best practices are produced are the interior of the organization.
An internal practice (or organization's practice) can become an accepted best practice. On the other hand, a best practice can be adopted and adjusted to an internal practice.

Data Understanding

2010-01-11T14:29:00.003-07:00

I have provided some formulas to facilitate presentation of thoughts around data management. I would like to add one to bring knowledge into the conversation. The formulas are as follows:
Data = Attribute + Value
Information = Question + Answer (Based on available data)
Knowledge = Information + Understanding
For instance, the attribute "amount" can be instantiated with the value "$1 million". Therefore, a data regarding a given amount is $1 million. To simplify the exercise, we will remove the context for now.
For the question, "What is the amount...? the answer could be "$1 million". This will provide some information.
Once understanding is added to the process, it will derive into knowledge that can be used to take some actions based on informed decisions.
Putting this example in the context of sales in a given fiscal quarter, it can be read as follows:
Data = "sales amount" + $1 million.
Information = "Total sales amount within the last fiscal quarter? + "$1 million"
Information on total sales amount within the last fiscal quarter can be enhanced by adding more information such as: this amount is lower than the previous fiscal quarter and lower than the same period in the last three fiscal years, etc. In addition, this amount did not reach the sales target for the quarter.
Analyzing this information it would be possible to generate some thoughts surrounding the data and consequently accumulate some knowledge regarding it.
Data understanding goes beyond collecting and organizing data. It requires transformation into information and then into knowledge. Here is where the value of Business Intelligence can be better appreciated.

Knowledge Assets

2010-01-08T13:10:00.004-07:00

Transformation of data into information is an in-progress activity in most organizations. However, transformation of information into knowledge is either in its infancy or has not started yet at most organizations.
Methods and techniques (e.g. data mining) for knowledge discovery has been available for several years. It has been the academia the one that has been most benefit from it. There are probably thousands publications of research results on this area of knowledge.
There are some steps to move towards the knowledge discovery realm:
1) Know your data. This implies identifying data assets (data that brings value to the organization), information needs, and then perform a gap analysis to determine which data and data sources should be integrated to satisfy such needs.
2) Manage your data. In a nutshell this can be broken down into having a BI strategy (which includes the DWH one), implementing the BI strategy (which includes DWH, reporting & analytics, and performance management solutions), having BI governance.
3) Learn from your data. Analytical applications support this step. Slicing, dicing, and drilling-through are important capabilities offered by these applications to support the process of learning and understanding your data.
In the exercise of understand data (or information), knowledge will be produced. This knowledge becomes your knowledge assets. So, here we are moving beyond data and information assets.
Knowledge assets are part of the intellectual property of the organization and represent a very important value add to any product or service the organization provides its customers with.
Now, these knowledge assets are nothing if the decision makers continue making decisions based on "gut feelings" (or common sense). Gut feelings do not require conscious thought. It is common to find decision makers providing "guesstimates" for a specific task.
In order to differentiate your organization from the competitors, it is crucial to accumulate and use knowledge assets and use them in the decision making process.
This could derive into patents, copyrights, trademarks, and publications. In other words, it is the intellectual property of your organization that is not only of great value but can be protected by laws.
Transforming information into knowledge requires a mental process. Understanding of information and application (practice or experience) results into expertise. Some of my formulas are:
Data = Value + Attribute
Information = Question + Answer
Knowledge = Information + Understanding
Your perception (feelings) of what you know determines what you do. This is actionable information or knowledge.
Using knowledge assets could help organizations improving their performance. It can be measured in terms of customer satisfaction, employee retention, etc.

BI Vision 20-20

2010-01-07T13:43:00.005-07:00

Open data will be more available from public and private sectors. This will change the conception of data architectures, information systems, and BI.
There will be stronger and better defined policies and more robust technology to facilitate data protection. This will represent new options for data integration approaches. This will increase the use of web services and alter the landscape for BI solutions.
There will be more focus on transforming data into knowledge.
Cloud computing is pushing a new way to architect data management solutions. Having data distributed in the cloud will move BI from in-house hosted solutions to open BI architecture with data elements and processes spread across the cyberspace.

Business Information

2009-12-17T09:09:00.008-07:00

Following my comments on Business Data, I would say that Business Information is information that is produced or consumed by organizations and is part of the business operations.

Business information is an organization's asset and should be managed as such.

Business Information Management

Let's analyze the basic functions plan, organize, direct, and control (PODC) for managing business information.

Plan: Before any planning activity regarding business information, it is important to determine what are the information needs. This can be done by following a business needs identification process that I will address later on. Meantime, let's focus on other aspects that contribute to a successful planning process in relation to business information management. These aspects include identifying from the business (or organization's) strategy: business goals, business objectives, processes and people involved in achieving these goals and objectives, and business drivers.
Organize: Once information needs have been identified and they are understood within the context of the organization (goals, objectives, drivers, etc.), information assets can be organized in a manner that they can contribute to the organization's strategy. Organizing the information assets also include identifying sources of data, processes to transform data into information, and information delivery.
Direct: After organizing information assets, it is important to have a strategic and tactical direction that will be followed to utilize these assets for improving business performance and minimize risk by making well-informed decisions. The function is used to provide direction in how to information assets are produced and consumed.
Control: Controlling, or coordinating, information assets, contributes to reach effectively and efficiently organization's goals and objectives. In other words, well-managed information become actionable information, information that is useful for making strategic decision. Then, it is when organizations clearly see the value of information and consider it an asset.

At this point it is easier to present the direct connection between business information management and business intelligence.

Information

2009-12-17T06:39:00.010-07:00

I submit a question to the business data post. The question is: What is the difference between data management and information management?

Well, I don't have a precise answer right now, but I would like to elaborate on this. I use to define information in the same manner I define data.

My formula to define information is as follows:

[ Information = Question + Answer ]context

Thus, answers to questions help people to "form" or "shape" ideas or concepts regarding a subject. In other words, answers to questions produce information regarding a subject. The process of forming or shaping ideas or concepts regarding a subject is the process that transforms data into information.

Quality of information depends on the quality of the question, the quality of the answer and putting these two together within the right context.

Information is obtained when an answer is provided to a specific question within a given context. Answers are based on data, which are also placed within a context.

BI Product Adoption Process

2009-12-16T21:34:00.000-07:00

Note: Based on Product Adoption Process Definition found at BusinessDictionary.com

BI Product Adoption Process

Five-stage mental process all prospective customers go through from learning of a new product to becoming loyal customers or rejecting it. These stages are

Awareness: prospects come to know about a product but lack sufficient information about it;
Interest: they try to get more information;
Evaluation: they consider whether the product is beneficial; BI Platform Evaluation.
Trial: they make the first purchase to determine its worth or usefulness; From Proof-of-Concept to Pilot Project.
Adoption/Rejection: they decide to adopt it, or look for something else.

Another explanation is that the customer moves from a cognitive state (being aware and informed) to the emotional state (liking and preference) and finally to behavioral or conative state (deciding and purchasing).

Two important aspects to consider before adopting a product are:

Product Maturity: The product is mature and well-established in the market;
Enterprise readiness: The enterprise is ready to take full advantage of the product.

A quick risk assessment can be derived from a combination of product maturity and enterprise readiness as described as follows:

Product Immature & Enterprise Not Ready = Failure can be anticipated
Product Immature & Enterprise Ready = High Risk
Product Mature & Enterprise Not Ready = Medium Risk
Product Mature & Enterprise Ready = Low or No Risk

Business Intelligence Strategy

2009-12-16T21:23:00.000-07:00

BI Strategy: Planning for Success

Step 1: Conceptual BI Architecture

Data Sources
Services
Presentation

Hint: "Use a building block approach"

Step 2: BI Roadmap

BI Framework
BI Initiatives
Sustainability

Hint: BI is neither a product nor a system

Step 3: BI Infrastructure

People
Processes
Technology

Hint: Maximize your existing infrastructure

Data Integration Architecture

2009-12-16T21:18:00.000-07:00

Hub-and-Spoke Data Integration Architecture

Extract from Data Sources
Pre-Processing
Data Integration Hub
Post-Processing
Load into Targets

Data Integration Process

2009-12-16T21:16:00.001-07:00

Note: Based on The Data Integration Process published at Microsoft TechNet

Step 1: Creating the staging database
Step 2: Synchronizing the staging database
Step 3: Populating the staging database
Step 4: Validating data in the staging database
Step 5: Loading data from staging to application database

Data Integration

2009-12-16T21:11:00.000-07:00

Heterogeneous Data Sources

Pulling together and reconciling dispersed data
Operational and analytic purposes
Maintained in multiple, heterogeneous systems

Extraction, Transformation, and Loading (ETL)

Data need to be

accessed and extracted,
moved and loaded,
validated and cleansed, and
standardized and transformed

Business Intelligence Platforms and Tools

2009-12-16T13:50:00.000-07:00

Business Intelligence Platform
A BI Platform is defined as a software that provides a set of common services to build, manage, and deploy scalable BI solutions.

A BI platform simplifies integration, analysis, and delivery through a number of BI capabilities
According to Gartner, BI platforms enable users to build applications that help organizations learn and understand their business

Business Intelligence Tool
A BI tool is a type of software designed to address specific BI capabilities. For instance, a BI reporting tool supports authoring, managing, and automated delivery of business reports.

BI tools can be used separately from BI platforms
BI tools can be integrated with different BI platforms

BI Platform Evaluation

There are different approaches to follow for evaluating BI platforms. My suggestion is to consider the following:

An up-to-date Magic Quadrant from Gartner. Bear in mind that the MQ indicates position of vendor and not specific products. The MQ provides a summary of strengths and cautions by vendor.

A BI Capability Matrix (BICM) from Gartner. It organizes 12 capabilities into three categories as follows:

Integration:
BI Infrastructure
Metadata Management
Development
Workflow and Collaboration
Analysis:
On-Line Analytical Processing (OLAP)
Advanced Visualization
Predictive Modeling and Data Mining
Scorecards
Information Delivery:
Reporting
Dashboards
Ad-hoc Query
Microsoft Office Integration

A product roadmap. Vendors or research firms (e.g. Research Directions) provide product roadmaps that are very insightful when making decision on which BI platform will be used in the coming years at your organization.

Technical publications such as blogs and white papers are sources of information about products, vendors and customer experience.

Proof-of-Concepts are valuable to get a better insight into the BI platform capabilities from a hands-on experience.

Comparison results from competitors are important but should be considered under the lenses of the BICM. Vendor criteria can be mapped into a unified matrix using the BICM.

BI Strategy is an important element to evaluat a BI platform. The BI strategy provide direction and guidance that is aligned with the organization's strategy and the IT strategy. It contains details on BI Framework, BI Architecture, etc.

As part of the evaluation, it is important to estimate a Total Cost of Ownership. It will provide decision makers with a more complete picture of the costs and benefits that could bring to the organization the BI platform.

To Take into Account

Market trends
Acquisitions
Early adopters (RDP)
Proof-of-Concepts (Accelerators)
Licensing Model
Product Roadmap
Training and Support
Consultants and Partners

Effective Business Analytics

2009-12-16T13:41:00.000-07:00

Business Intelligence (BI) creates a learning environment that enables organizations to apply more effectively business analytics. BI transforms data into information and then into knowledge by:

Integrating data from heterogeneous sources

Consolidating, organizing and standardizing data in repositories such as data warehouses and data marts

Providing tools that facilitate anaytical processing

Enabling business users to access and visualize multidimensional data

Displaying complex calculations and key performance indicators through integrated portals

Supporting processes for authoring, managing and delivering reports

Applying data mining techniques for uncovering patterns, forecast data and potentially make predictions.

Business Intelligence

2009-12-09T00:16:00.001-07:00

Business Intelligence involves people, processes and technology. From the definition given by David Loshin in his book "Business Intelligence: The Savvy Manager's Guide", Business Intelligence turns data into information, information into knowledge, and knowledge into plans that drive profitable business actions.

BI minimizes risk by supporting well-informed decisions providing accurate and timely information to the right people in the format they needed.

BI technology facilitates integrated monitoring, analysis and planning for optimized organization’s performance.

Areas of Business Intelligence

Data Warehousing
Report & Analysis (Business Analytics)
Performance Management

Business Data

2009-05-01T16:11:00.002-06:00

One of the purposes of creating this blog is to discuss data inside and out. Paraphrasing this, I should say, to discuss about data from Information Management (IM) perspective, i.e. from a technical (IT) and non-technical (business) perspective.

Actually, I have more interest in looking at data as "business data" rather than just writing a lot of stuff that can be found in any IT book.

Given my background, it would be difficult to avoid making technical comments. Indeed, it is better to provide those comments, as they are the ones that make my contributions valuable in terms of data management.

Let's start a see how things progress...

Business data is data that is produced or consumed by organizations and are part of the business operations. It is not a new type of data. My preference for adding the qualifier "business" is to emphasize that it is owned by the business. The expectations is that by calling it, business data, it will raise awareness that data is not something that only IT should care about.

Then, the first step is to establish the appropriate ownership. The next step is to recognize the value of data to the organization. Business data is an organization asset and should be managed as such.

The third step is to manage those data assets. Finding out where data assets are produced, how they are transformed, moved, copied, stored, etc. sounds like a lot of work. Here is when data management starts playing an important role to support organizations.

What is data management? and how does data management support organizations?

This will be the subject of my next blogs. For now let's summarize:
1. Business data is data used by organizations in their day-to-day operations.
2. Business data are organization's assets.
3. Data assets should be managed.

Data Element

2009-04-28T09:12:00.000-06:00

A data element is the lowest level of granularity or detail in the Universe of Data. It is an atomic unit of data that has precise semantics. Data elements cannot be subdivided, they are the most elementary units of data that can be defined in a dictionary or repository.

Data

A formula to define data is as follows:

[ Data = Attribute + Value ]context

Data is the association of a value to an attribute within a given context.

For example, the value 10 does not represent data itself. It can be anything. Let's say, it can represent number "ten" in the decimal numerical system, or number "two" in the binary system. Once this value is associated to an attribute, e.g. Number-of-Cities, a meaning is activated and data appears.

The association "Number-of-Cities + 10" in the context of the Roman Empire is also known as Decapolis. Now, this data is raw material that can be used to build some information within this context.